Amps
[Top] [All Lists]

[Amps] Linear amplifier tuning

To: <amps@contesting.com>
Subject: [Amps] Linear amplifier tuning
From: "Jim Thomson" <jim.thom@telus.net>
Date: Wed, 21 Dec 2011 08:52:04 -0800
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Date: Tue, 20 Dec 2011 19:20:48 -0600
From: Edwin Karl <edk0kl@centurytel.net>
Subject: [Amps] Linear amplifier tuning

The discussion about tuning would not happen if we all still used 
o'scopes with a
trapezoid pattern. On such a pattern you can clearly see "flat topping" 
when the output
loading cap is too tight, a slight reduction in capacity results in a 
clear, sharp point
at the end of the triangle.

Additional tuning just upsets the whole thing. While an out put meter 
might again
see a small increase in power, the additional 10-20 watts won't make a 
difference,
but the splatter from your flat topping amplifier will.

The aforementioned 5% change in output loading is a good rule of thumb, 
lacking
a 'scope

Merry Christmas to all!

ed K0KL

##  You can easily achieve the same effect with any good pep wattmeter, like 
either
a Coaxial dynamics or an array solutions  'power master'.   Just alternate 
between the
tune and load cap  a few times, till u have max po one the wattmeter.  Then if 
the
load control is cranked a bit more UN-meshed, till the po  drops 2%, ur done. 
All u are doing by UN meshing the load cap is  you are setting it where it 
would be , IF peaked for
even more PO.  

## another method, that amounts to the same thing..is to just use slightly more 
drive, then
peak the amp up for say 1.6 kw pep out, instead of 1.5 kw pep out.   Then 
switch to 
ssb/cw, etc... and reduce the po from the xcvr slightly, such that pep out on 
the pep meter
is  now 1.5 kw pep.  Then it's linear, plus it will now handle any overshoots 
from the xcvr.

##  i just pulse tune my amps anyway... with 60 wpm dots.  On the  yaesu MK-V, 
in the
menu, you can shorten/lengthen  both dots and dashs, individually.  Default is 
dots= 10
and dashes = 30.   For normal CW use, I use  dots=10... and dashes =36. 

##  For pulse tuning on CW,   I reduce the duration of the dots...down to dots 
= 7
With dots = 10.... duty cycle is 50%  on the plate current meter.   By varying 
the
dot duration,  from  1-10 in the menu,   the  duty cycle can be varied from  
10-50%.
At dots= 7, it's about 30%..which is just fine. 

## a trapezoid pattern involves sampling 1st the xcvr's output... and also the 
amps output,
and that will work too.  But if just sampling the amps output only [xmas tree 
pattern on ssb],
ur wasting ur time, since it won't show flat topping unless it's really gross 
flat topping. 

##  If u have flat topping, u don't have the amp tune dup for max po  for XXX  
amount of 
drive anyway. 

## carl is correct. depending on loaded tank Q  on a band per band basis, you 
may well
have to tweak the TUNE control dead last.   It's  Ok  to have the LOAD control 
slightly
more UN-meshed, but it's a disaster if the  TUNE control is off resonance.  On 
my drake L4B amps,
on 80m, the tune control is tweaked for max po with just 25 w of drive.  After 
that, the tune control
doesn't get touched, just the load control...and that's at 800 ma, and more 
drive of course.  On most amps
the TUNE control needs tweaking as drive levels increase. 

##  It makes life a lot easier if you replace the knobs etc,  on your  tune + 
load control..and just replace
em both with jackson bro's  6:1 redux  ball drives.  Then you have big diam 
skirts, which are calibrated
from 0-100  across  180 degs.   Then it's silky smooth tune and load 
adjustments.  Then just log the
tune + load settings  for  cw/ssb  for  each band.   After that, it's  just ..  
"dial it up by the numbers"  

##  with vac tune and load caps, it gets even better, since most are like 14-24 
full revolutions from
min to max C.   With the old  Groth type  turns counters, they not only count 
turns, they also  calibrate
the 3"  diam skirt  from 0-100.  A 3"  diam skirt has a 9+ inch circumference.  
A 500 pf tune cap is aprx
18 turns from min to max....or aprx   28-30 pf per knob revolution.  OK, now 
divide 28-30 pf  by 100
x small increments on the 3"  skirt..amd each increment is now just .3 PF  !!   
The same groth can also be used
on any roller coil.  Most other's  will  split each rev on their turns counter 
down to tenths.  The frontier engineering
turns counters are digital readout, and also count down to 1/10ths of a rev. 
http://www.freng.com/   [$120.00 each]

## as long as the SWR hasn't changed much,  dialing up.."by the numbers"  makes 
life a lot easier..whether air caps
or vac caps are used.  BTW, jackson bro's in the uk has just informed us that 
they will again be making their various
6:1 ball drives, and plane 6:1 drives only...after xmas.   They are things of 
beauty. http://www.mainlinegroup.co.uk/jacksonbrothers/index.htm

later...... Jim   VE7RF 


_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>